Inventors’ mobility and the geographies of innovation: the case of mRNA COVID-19 vaccines

Recent evidence shows that inventors’ mobility is a primary channel for, on the one hand, facilitating the spatial diffusion of knowledge—often tacit and sticky—and, on the other, diversifying regional innovation assets by escaping cognitive lock-ins and enabling technological change. This distinctive role has drawn greater attention to the phenomenon, particularly from a geographical perspective, as it exemplifies the nexus between spatiality and innovation: inventors’ mobility not only contributes to the recombination and reuse of knowledge vital to the innovation process, but also shapes the geography of innovation and is therefore among the main drivers of pronounced spatial inequality in the allocation of resources, increasingly concentrated in a few hubs. Understanding this phenomenon is thus crucial for its impact on regional innovative capacities and, consequently, for equitable and effective regional development. The aim of this thesis is to advance understanding of the spatiality of inventors’ mobility by identifying the spatial structures that emerge from it, what these structures imply, how they change and co-evolve with the innovation process over time, which factors influence mobility, and how mobility is perceived by inventors themselves. The principal case study reconstructs the innovation process that led to the implementation of mRNA vaccines—an innovation that has transformed modes of producing preventive medicines and enabled a rapid response to the COVID-19 pandemic. This process is particularly suited to the thesis because it spans a relatively long period during which research was initially driven by basic science—primarily involving public institutions, universities, and start-ups, in contrast to the skepticism of large firms—before a reversal in industry attitudes following breakthroughs that made the vaccine economically viable and scalable. Tracing inventors’ mobility along this process sheds light on the spatial dynamics it produces. Methodologically, the thesis adopts a multi-method and multi-level design. A relational-quantitative analysis using network methods explores complex and emergent spatial structures, alongside a qualitative analysis oriented to inventors’ subjective perspectives through their life histories. The discussion begins by retracing the main phases of the innovation process that led to mRNA vaccines, highlighting that for a long period it was driven by basic research—highly uncertain and at times marginalized. Together with the vaccine’s distinctive features, this meant that for the first decades research proceeded across multiple disciplines and places, until the recent acceleration brought about by the concrete possibility of developing the product. Subsequently, the economic-geography literature on inventors’ mobility is examined with a critical lens, bringing to light three main areas of concern: (i) divergences over mobility’s role as a source of knowledge diffusion and regional diversification versus evidence of concentration and specialization; (ii) ambiguities regarding the territorial effects of mobility, especially for origin regions; and (iii) a limited understanding of the drivers of mobility, which is predominantly framed within a microeconomic approach. To advance understanding of these issues, three empirical studies are presented. The first study identifies the spatial structures emerging from inventors’ mobility and their implications for stability, efficiency, and equity. I map the career trajectories of the principal vaccine inventors to reconstruct a directed city-to-city mobility network. The analysis shows a scale-free network, i.e., a complex network with specific properties. Community detection reveals a core-periphery structure in which technology hubs form much denser inter-city connections that facilitate knowledge exchange and diversification. A distinctive feature of hubs thus emerges: rather than merely accumulating human capital, they promote its transit and circulation. Exponential Random Graph Models (ERGMs) estimates corroborate these findings, highlighting also the predominance of intra-national mobility and the importance of maintaining a balanced mix of education-oriented and employment-oriented institutions to avoid losing human capital. In terms of implications, these geographical configurations prove highly resilient to random shocks that stress the ecosystem, yet strongly vulnerable when shocks strike the hubs, rendering the entire innovation system dependent on them. Moreover, network topology is not neutral with respect to inventors’ mobility; it can itself amplify centralization, thereby posing risks to the equity and efficiency of the innovation process. The second study complements the first by moving beyond a static view of the geography of innovation. Incorporating time, it examines how mobility-generated spatial structures change and co-evolve with stages of the innovation process. I reconstruct the mobility of all inventors involved and build a temporal network linking cities via inventors’ moves. The process is divided into two periods: an initial phase characterized by low technological opportunities and even lower innovative opportunities—dominated by basic, exploratory research—and a subsequent phase in which innovative opportunities surpass technological ones, marking readiness for production and market introduction. The analysis indicates that in the first period the mobility-generated structures are quasi-random, non-hierarchical, and the network is sparse and fragmented. In the second period, the network becomes centralized around a few cities that connect a large share of the system. Temporal ERGMs (TERGMs) estimates show that in the first period inventors tend to move toward cities that have produced knowledge—especially incremental knowledge—whereas in the second period knowledge plays a smaller role, and inventors gravitate toward hubs holding complementary assets necessary for scaling and commercialization. The third study adopts a biographical approach to gain a deeper understanding of inventors’ motivations and viewpoints. It examines which factors researchers deem most important in their career and location choices, and how they perceive the spatial dynamics produced. Analysis of eleven life stories reveals an interplay between subjective drives and objective contexts shaping inventors’ decision-making. The primary engine is an intrinsic, indeterminate “blind” motivation—often described as curiosity, the desire to learn and discover, and, given the field, to care. This initial impulse meets the socio-spatial context, which filters and channels it into concrete choices. Among the objective factors that inventors consider crucial are access to funding—mainly to sustain uncertain research rather than to maximize earnings—and the need to establish and preserve social ties. This dyadic dynamic between the subjective and the objective is perceived as co-evolutionary: researchers’ choices and work influence socio-spatial contexts, which in turn reshape decision-making process. A perceived geographical division of the innovation process emerges; whereby certain places are more oriented toward particular phases due to this co-evolution reinforcing territorial specialization. At the same time, inventors express concerns that market tendencies to favor some phases over others risk marginalizing the latter. This thesis highlights the interdependent relationship between the complex geographies generated by inventors’ mobility and the innovation process. In particular, it shows how these geographies make the innovation system highly dependent on a small set of hub cities which, while crucial for the circulation of inventors and knowledge, simultaneously render the system vulnerable to shocks that may strike them, given their capacity to connect the entire spatial system. The thesis also shows that hubs do not play the same role throughout the innovation process: their influence is pivotal mainly in its final stage, whereas in the early stages inventors’ mobility is primarily driven by the search for knowledge, producing an evolution of spatial patterns from quasi-random to centralized. Finally, the thesis underscores the dialectical, co-evolutionary relationship between inventors’ subjective motivations and the socio-spatial contexts they help create, and it reports their concerns regarding the problems this process poses for the production of novel knowledge. The analysis points to the need for policies that mitigate the process’s distortive and unequal effects by fostering the circulation—rather than the retention—of human capital, and by directing efforts toward research that remains capable of taking the risks required to sustain it.